Evidence of Children’s
Achievement
IPR experts hold policy briefing on Capitol
Hill
Fall 2006, Volume 28, Number 1
 |
| Thomas
Cook takes a question on pre-K programs while Larry Hedges listens. |
On May 19 on Capitol Hill, three national experts reviewed the latest evidence indicating how teachers, pre-K programs, and economic policies could boost the success of children in America’s classrooms.
IPR Faculty Fellows Greg J. Duncan, Thomas D. Cook, and Larry V. Hedges offered the crowd of more than 100 people—including policymakers, journalists, academics, and advocates—policy suggestions rooted in evidence-based studies. The Joyce Foundation provided funding for the policy briefing.
Teacher Effectiveness
Anecdotes abound of how great teachers have shaped their students’ lives. But most parents seem more concerned about which school their child attends, noted educational researcher Larry V. Hedges, Board of Trustees Professor in Statistics and Social Policy. So parents take great pains to buy houses in the “right” school district or pay tuition for their children to attend the “right” school.
 |
|
Larry
Hedges makes a point about teacher effectiveness. |
According to Hedges’ research, however, teachers can have a larger effect on student achievement than schools. He found that the most reliable studies of teacher effectiveness use randomized designs. Yet only one study so far has used random assignment to measure it: Tennessee’s Student Teacher Achievement Ratio (STAR), a $12 million, four-year study that covered 11,000 students in 79 schools from kindergarten to 3rd grade.
Analysis by Hedges and his team revealed that teachers account for 7 percent of the variation in reading achievement and 12 percent in math. These results suggest that a child moved from a below-average teacher (ranked in the 25th percentile) to an above-average teacher (in the 75th percentile), would experience significant gains in reading and math. These gains double if the child is then moved to an excellent teacher (in the 90th percentile).
Teacher effects, Hedges noted, are three times bigger for schools serving mostly poor students than for those schools serving mostly wealthy ones. “Thus, it matters more which teacher you happen to get in a poor school than in a rich school,” he said.
To improve student achievement, he suggested policymakers craft programs based on teacher choice, teacher accountability, and proven approaches to effective teacher development. Such progams have the potential to increase student achievement as much as, or more than, reducing class size drastically and at less cost. Such a strategy could have a larger “bang for the buck,” especially in the nation’s poorest schools, he said.
Pre-K: State Programs and Head Start
Preschool programs do make a difference—at least in the short-term—and merit being expanded, noted social psychologist Thomas D. Cook, Joan and Sarepta Harrison Chair in Ethics and Justice. Across the nation, 68 percent of 4-year-olds and 39 percent of 3-year-olds are enrolled in some type of preschool. But which type of preschool is better and for whom?
Cook observed disproportionate racial representation between public and private preschool enrollment. More than 70 percent of 4-year-old black and Hispanic children attend public programs, contrasted with 43 percent of white children. Additionally, he noted that black 4-year-olds are over-represented in Head Start programs, 28.7 percent versus 4.4 percent for whites and 11.8 percent for Hispanics.
Recently, discussions have taken place to roll federal Head Start monies into state block grants. One source of evidence adduced to support this policy comes from a recent National Institute for Early Education Research (NIEER) study. Using the study, some have argued that state programs have larger achievement effects on preschoolers than does Head Start, as determined from Westat’s First Year Head Start results.
 |
| An audience member asks about methods to
|
“But you have to ask yourself, ‘Are these evaluations similar?’” he cautioned. “And the answer is ‘no.’”
While the Head Start study uses a nationally representative sample, NIEER only looked at five of the most well-established state preschool programs—with four ranking higher than the national average. Other differences included: Head Start children were poorer than children in the state-run programs; the Head Start study was more comprehensive, measuring a child’s socio-emotional state and health—not just cognitive achievement; and the control group used to evaluate Head Start had a higher percentage of children in it who were in an alternative preschool program, thus creating a higher threshold for Head Start to reach in order to declare it effective.
No solid evidence backs the claim that state pre-K programs are better than Head Start programs at raising children’s achievement, he said. “These decisions about block grants may have to be made,” he continued. “But they are likely to be made on political grounds that are less informed by science than one would hope.”
Economic Policies and Achievement
Given that the black-white achievement gap between children entering kindergarten is about 8 points on an IQ-type scale, economist Greg J. Duncan asked the audience to imagine what kind of a program could show a difference of about 4 points between an experimental and a control group.
Interestingly, it was not an educational program design but an economic one—Milwaukee’s New Hope Project—that generated this significant difference, Duncan pointed out. New Hope was a voluntary, work-based, antipoverty initiative that offered participants varying packages of job search assistance, earnings supplements, subsidized health insurance, and/or subsidized child care.
Duncan, Edwina S. Tarry Professor of Education and Social Policy, examined 11 randomized studies from the Next Generation Project, in addition to the Earned Income Tax Credit (EITC), to see how they affected children’s outcomes.
Overall, he found that the earnings supplement programs seem to have the most consistently positive impacts on achievement for children up to 5 years old, with positive gains across all of these programs. In contrast, only one of the 11 programs using either human-capital development or work-first as a treatment registered positive increases in children’s achievement.
The one puzzle was the case of preteens and adolescents, who were somewhat negatively affected in all of the studies. Duncan explained the reason might be that working mothers have less time to supervise their adolescents’ behavior and often leave them to care for their younger brothers and sisters. “All of the negative impacts show up among the adolescents who had younger siblings,” he said.
“These studies show that the design of welfare reform and tax policies can indeed affect kids’ achievement,” he said. “They should be included alongside preschool and school-based policies.”
IPR Faculty Fellow Therese McGuire said, “With the bulk of attention in the educational sector focused on No Child Left Behind and what schools are doing, it is refreshing to see these thorough, evidence-based perspectives on how to improve the educational outcome of America’s children through teachers, preschools, and economic policies.” She is Beatrice Foods Research Professor of Strategic Management in Kellogg and director of the policy briefing series.
Click here to view the video or the slide presentations.
Click here for more information about the IPR faculty presenters and their research.